*3.6. Varicocele*

Varicoceles are a common cause of infertility, found in at least 5% of men with NOA, and the most common cause of secondary infertility [47,48]. Varicoceles are dilated veins (spermatic and pampiniform plexus veins) within the spermatic cord and may result in subsequent testicular dysfunction [49]. Different mechanisms of testicular dysfunction in men with varicoceles have been proposed, including testicular parenchymal hyperthermia, blood-testis barrier dysfunction, and testicular hypoxia [50]. The ultimate downstream effect is the generation of reactive oxygen species and damage to testicular cells [50–52]. Damage to Sertoli, Leydig, and germ cells can result in abnormal or decreased sperm production, as well as deficient testosterone production [49,53]. Although varicoceles clearly contribute to testicular dysfunction, a varicocele is unlikely to be the primary cause of NOA since only 5–10% of men with NOA and clinical varicocele will have enough sperm return to the ejaculate after varicocele repair to avoid testicular sperm extraction [54].

#### *3.7. Other Causes of NOA*

NOA may also be acquired for a wide variety of reasons including genitourinary infections, such as post-pubertal mumps orchitis, or various classes of medications. Common medication categories with documented negative impacts to fertility include exogenous testosterone or other androgen-modulating medications, psychiatric medications, and antihypertensive medications, which all have been documented to modulate the hormonal environment resulting in decreased or absent sperm production [55].

#### **4. Optimization of Sperm Production Prior to Surgical Retrieval**

Various abnormalities, including hormonal deficiencies and testicular dysfunction, may contribute to abnormal spermatogenesis and decreased or absent sperm production. The production of sperm requires adequate levels of (serum and intratesticular) testosterone, and the goal of optimization is to increase testosterone levels. Serum testosterone levels can be optimized prior to surgical sperm extraction by administration of hormone analogs and modulators, including gonadotropins, aromatase inhibitors (AI), and selective estrogen receptor modulators (SERM) (Table 2). Unfortunately, there is no high-level evi-

dence to support use of medical therapy prior to sperm retrieval, despite many anecdotal applications of medical therapy prior to attempted sperm retrieval for NOA [22]. Certainly, it is conceptually appropriate to treat low testosterone levels in men with NOA. However, retrospective data from a large series of men with NOA sugges<sup>t</sup> that the benefits of treatment are quite limited, with men receiving medical therapy to raise testosterone levels having SRR of 51% (151/307) compared with men not receiving medical therapy prior to surgical intervention who had SRR of 61% (25/41) (*p* = 0.31) [56].


**Table 2.** Medical therapy for hormonal optimization prior to sperm retrieval.

AI, aromatase inhibitor, SERM, selective estrogen receptor modulator; rhFSH, recombinant human folliclestimulating hormone; hCG, human chorionic gonadotropin; IU, international unit.

Medical therapy used to optimize testosterone levels aims to increase testosterone production and decrease estradiol levels. Elevated estradiol levels, typically greater than 60 pg/mL, can suppress hypothalamic gonadotropin secretion and subsequently inhibit testosterone production [6]. Testosterone-to-estradiol (T:E) ratios are normally greater than 10, and fertile men have a mean T:E ratio of approximately 15 (14.5 ± 1.2) [57,58]. Men with infertility have lower T:E ratios, with NOA men typically in the range of 7 (6.9 ± 0.6) and men with KS approximately 5 (4.4 ± 0.5) [58–60]. Gonadotropins can stimulate testosterone production and spermatogenesis in men with low gonadotropin levels secondary to congenital or acquired disorders [60]. hCG may be used as an luteinizing hormone (LH) substitute alone or in combination with an FSH analog (recombinant human FSH (rhFSH) or human menopausal gonadotropin (hMG)) to stimulate testis growth, testosterone production, and spermatogenesis [60]. AIs, including anastrozole and letrozole, prevent the actions of aromatase, which is present in peripheral tissues, in converting testosterone to estrogen. AIs have been shown to be effective medical therapy to increase SRR in men with KS and in infertile, non-KS men with abnormal T:E ratios [61,62]. SERMs, such as clomiphene citrate and tamoxifen, provide benefits by inhibiting the negative feedback exerted on the hypothalamic-pituitary-testis (HPT) axis by estrogen. With decreased negative inhibition, higher levels of LH and FSH can be achieved, resulting in increased serum testosterone levels and improved sperm production. Several studies have demonstrated positive impacts on semen parameters in infertile men taking clomiphene citrate [63,64].

Repair of clinical varicoceles has been demonstrated to improve serum testosterone levels, as well as spermatogenesis in men with oligozoospermia [48]. After varicocele repair in men with NOA, the potential improvement of spermatogenesis may result in enhanced SRR, although there is no high-level evidence to support such intervention [54].

#### **5. Intracytoplasmic Sperm Injection for NOA**

Prior to the advent of ICSI, men with NOA had no means for conceiving biological children. ICSI was first introduced in 1992, and three years later, in 1995, sperm retrieved from an NOA patient was used successfully with ICSI [65,66]. Since the development of ICSI, numerous studies have been performed to examine factors that may predict or be associated with increased ICSI success.

#### *5.1. Predictors of ICSI Outcomes*

Limited pre-operative variables exist which predict success of SRR and ICSI. For predictors of SRR, patient age, serum hormone levels, and testicle size have been evaluated,

however, conclusive evidence is lacking that any of these factors is predictive of successful sperm retrieval [67–69]. Testicular histopathology does provide some prognostic information for SRR, but it is not routinely recommended for diagnosis of NOA, as the diagnosis can be made clinically based on FSH > 7.6 and testis length < 4.5 cm in about 90% of men with this condition [22,68,70,71]. Similarly, no clinical or biochemical factors have been found to be predictive of ICSI outcomes [3]. Additionally, testis histology has not been shown to significantly influence clinical outcomes after ICSI [71]. There may, however, be an association of the number of sperm found at time of surgical retrieval with the number of clinical pregnancies [72]. Further work is needed to determine if any preoperative factors, in conjunction with female factors, can predict ICSI outcomes.

#### *5.2. ICSI Outcomes in NOA Men*

Understanding clinical outcomes after ICSI are important when counseling men with NOA and their partners prior to surgical sperm retrieval. A study derived from the National Assisted Reproductive Technology Surveillance System (NASS) found that men with infertility (including non-azoospermic and azoospermic men) had a clinical pregnancy rate (CPR) of 48% and live birth rate (LBR) of 40%, which was similar to rates in men with no infertility (CPR 44.9%, LBR 36.5%) [73]. A clear limitation of this study was that the various etiologies of male infertility were not specified or separately examined and thus, the CPR and LBR may not hold true for all etiologies of NOA. Although NOA men represent the most severe phenotype of those with male infertility, the majority of studies, similar to that previously described, have pooled men with NOA regardless of etiology which limits the overall generalizability of the data. A comprehensive summary of SRR, biochemical pregnancy rate (BCPR), CPR, and LBR from studies investigating ICSI outcomes between 1997 and 2020 in NOA men is presented in Table 3.


**Table 3.** Studies reporting intracytoplasmic sperm injection outcomes in men with non-obstructive azoospermia.


**Table 3.** *Cont.*


**Table 3.** *Cont.*


**Table 3.** *Cont.*

AZFRc, azoospermia factor region deletion in locus c; SRR, sperm retrieval rate; BCPR, biochemical pregnancy rate (elevated serum hCG); CPR, clinical pregnancy rate (heartbeat or gestational sac detectable by ultrasound); LBR, live birth rate; MR, miscarriage rate; NOA, non-obstructive azoospermia; NR, not reported; TESA, testicular sperm aspiration; cTESE, conventional TESE; mTESE, microdissection TESE; TESE \*—type of TESE not specified; €, ectopic pregnancy rate; KS, Klinefelter syndrome; YCMD, Y-chromosome microdeletion; RT, radiation therapy; MESA, microsurgical epiddiymal sperm aspiration. "a" and "b" were used to denote different patient cohorts examined within one study.

> A recent meta-analysis examining sperm retrieval as well as pregnancy and LBRs was performed [3]. This review compared SRR after conventional TESE (cTESE) with that after microTESE, and found that the per procedure SRR was 45–49%, and was not able to identify differences between conventional or microsurgical methods of sperm retrieval because the included studies did not include comparator trials [3]. Meta-regression analysis further demonstrated that SRR was independent of both age and hormonal parameters [3]. Testis volume greater than 12.5 mL was found to be associated with a greater than 60% chance of successful sperm retrieval with an accuracy of 86.2% [3]. The BCPR (diagnosed by positive serum hCG in the female partner) was 25–32% per ICSI cycle, and LBR was 20–28% [3]. It is important to note that the patient cohorts included in this meta-analysis were heterogenous with varying NOA etiologies, making the comparison of outcomes between cTESE and mTESE less valid. A previous meta-analysis which included fifteen comparative studies demonstrated a 17% higher likelihood of sperm retrieval success when performing mTESE compared to cTESE [2]. Additionally, it was noted that men who underwent mTESE had failed prior cTESE or TESA, which also may have underestimated the increase in sperm retrieval rate with mTESE [2]. Several, smaller studies have been performed examining NOA men based on underlying etiology, including KS, YCMD, malignancy, and cryptorchidism, and these will be discussed further.

## 5.2.1. Klinefelter Syndrome

A meta-analysis of 37 studies found a cumulative SRR of 44% (39–48%) per TESE procedure in KS patients, with no significant difference between cTESE and mTESE [114]. ICSI outcomes were available for 29 of the 37 studies in the meta-analysis, and reported a cumulative CPR, defined by ultrasound detection of a gestational sac or heartbeat, of 43% (36–50%), and LBR of 43% (34–53%) per ICSI cycle [114]. SRR, CPR, and LBR in this analysis were independent of patient age at time of retrieval as well as testis volume, and serum hormone parameters [114]. Additionally, no differences between use of fresh versus frozen sperm were observed [114]. Again, it is important to note that this meta-analysis examined studies where the patient cohorts were not entirely made up of KS patients. In one of the largest published studies on SRR in KS patients, we report a SRR of 66% (Table 4) [61]. In our experience with KS patients, the appearance of tubules within the testis is unique amongs<sup>t</sup> men with NOA. Instead of typically having sperm production throughout an individual seminiferous tubule, KS patients tend to have focal enlargement of otherwise sclerotic tubules within the testes. This appearance requires an intensive search within

these typically very atrophic testes to find the millimeter-sized segments of tubules that may contain sperm. In addition, the number of sperm retrieved tends to be so small that sperm are typically not able to be frozen for later use. Therefore, the numbers of sperm obtained may be only adequate to inject available oocytes during a programmed, fresh in vitro fertilization (IVF) cycle.


**Table 4.** Sperm retrieval rates in non-obstructive azoospermia by etiology.

P.N.S., Peter N. Schlegel, attending urologist at Weill Cornell Medicine.

#### 5.2.2. Y-Chromosome Microdeletions

Little is known regarding ICSI and clinical outcomes in NOA men with YCMD given that many of these studies excluded men with YCMD or included men with YCMD in a larger cohort of azoospermic men (Tables 3 and 4). SRRs differ drastically depending on the site of microdeletion. Sperm can be surgically retrieved in up to 70% of men with AZFc deletions and a subset may have low concentrations of sperm in the ejaculate, whereas no reports of sperm retrieval in men with complete AZFa or AZFb deletions have been effectively documented [17,116]. One study examining ICSI outcomes using ejaculated sperm demonstrated no significant difference in pregnancy, live birth, and miscarriage rates in men with AZFc microdeletions compared to those with other sources of infertility and no evidence of YCMD [117]. Another study found that men with AZFc microdeletions had a significantly increased fertilization rate when ejaculated sperm was used compared to testicular sperm [118]. With ejaculated sperm, pregnancy rate was 47% compared to 14% with testicular sperm [118]. Unfortunately, no predictors of successful sperm retrieval have been identified in this cohort, and it is important to inform couples that any male offspring will harbor the same Y-chromosome mutations [118,119].
